Knowledge-based identification of artery branches in cine-angiograms
Saburo Tsuji, Hidetoshi Nakano · International Joint Conference on Artificial Intelligence · 1981
This paper describes an image understanding system which interprets line segments detected In cine-angiograms, cine-film records of radiographic images of coronary arteries, into a model of the arterial system based on expert knowledge on the artery. A production system accesses rules on shapes and structures of the blood vessels to evaluate the geometrical parameters of sequences of segments, and a tree of hypotheses on the Identities of the individual segments in each view Is generated. Finally, the hypotheses obtained from different views are integrated to find a more reliable Interpretation of the images.